[J30] Reconfigurable Resistive Switching Memory for Telegraph Code Sensing and Recognizing Reservoir Computing Systems

Abstract

Reservoir computing (RC) system is based upon the reservoir layer, which non-linearly transforms input signals, typically in the time domain, into high-dimensional states, facilitating simple and rapid training in the readout layer—a linear neural network. These layers require different types of devices—the former demonstrated as diffusive memristors and the latter prepared as drift memristors. The integration of these components can increase the structural complexity of RC system. Here, we demonstrate a reconfigurable resistive switching memory (RSM) capable of implementing both diffusive and drift dynamics in a single device. This reconfigurability is achieved by preparing a medium with a three-dimensional (3D) ion transport channel (ITC), enabling precise control over the robustness of the metal filament that determines diffusive or drift memristor operation. The 3D ITC-RSM operates in a volatile threshold switching (TS) mode under a weak electric field and exhibits various short-term dynamics (non-linear variation, echo-state property, and separability) that are confirmed to be applicable as reservoir elements in RC systems for temporal information processing. Meanwhile, the 3D ITC-RSM operates in a non-volatile bipolar switching (BS) mode under a strong electric field, and the conductance modulation metrics (excitatory post-synaptic current [EPSC], paired-pulse facilitation [PPF], long-term potentiation/depression [LTP/LTD]) forming the basis of synaptic weight update are validated from non-volatile 3D ITC-RSM, which can be utilized as readout elements in the readout layer. Finally, an RC system integrating the reservoir and readout layers is designed for the application of reconfigurable 3D ITC-RSM. The proposed RC system performs real-time recognition on Morse code datasets, widely known among telegraph codes, achieving recognition rates over 98%, even for highly similar code signals.